Weaviate
Weaviate is an open-source vector database for semantic search, RAG and AI memory, available self-hosted or as a managed cloud service.
About this tool
Weaviate is a vector database that stores both data objects and their vector embeddings, so applications can search by meaning rather than exact keywords. It combines vector similarity with keyword filtering, retrieval-augmented generation and reranking in one query interface, and it is used as the foundation for semantic search, chat over documents and AI agents.
Vectors can be created automatically at import through integrated model providers such as OpenAI, Cohere and Hugging Face, or supplied pre-computed. Production features include multi-tenancy, replication and role-based access control. Weaviate is open source and can be self-hosted, while Weaviate Cloud offers a free sandbox, usage-based plans and dedicated deployments on AWS, Google Cloud and Azure.
What you can do
Hybrid search
Combine vector similarity with keyword filters, generation and reranking in a single query.
Built-in vectorisation
Generate embeddings at import with integrated model providers, without a separate pipeline.
Query Agent
Ask questions in natural language and let the agent translate them into database queries.
Multi-tenancy and RBAC
Isolate data for many customers and control access with roles in production deployments.
Who is it for?
- A SaaS company adds semantic search across each customer’s documents with tenant isolation.
- A developer builds a RAG chatbot over product documentation.
- An e-commerce site recommends similar products using vector similarity.
Before you choose
Self-hosting requires managing memory-intensive indexes; review vector index types and compression options to control costs.
Uptime commitments and support response times differ between cloud plans, so match the plan to your production needs.
Frequently asked questions
Is Weaviate open source?
Yes. Weaviate’s core database is open source on GitHub and can be self-hosted, alongside the managed Weaviate Cloud service.
Is there a free Weaviate Cloud plan?
Yes. A free sandbox includes one cluster with up to 100,000 objects, plus limited embedding and Query Agent requests.
Do I need a separate embedding service?
No. Weaviate can vectorise data at import using integrated providers, or you can import your own pre-computed embeddings.
Check the official website for current features, pricing, and terms.
Similar tools
Pinecone
CodingPinecone is a managed vector database and retrieval platform for AI apps and agents, combining semantic, full-text and hybrid search in one index.
Chroma
CodingChroma is an open-source search database for AI, offering vector, full-text and metadata search, with a serverless Chroma Cloud and bring-your-own-cloud options.
Cohere
CodingCohere is an enterprise AI company offering Command language models, Embed and Rerank for search, and North, an agentic platform deployable in private environments.
Baseplate AI
AutomationBaseplate AI is a development‑focused platform that lets engineers embed data, run hybrid searches, collect user feedback, and launch AI‑driven applications.
LlamaIndex
CodingLlamaIndex is known for its open-source RAG and agent framework and now focuses on LlamaParse, a platform for parsing, extracting and indexing documents for AI.
Amazon Bedrock
CodingAmazon Bedrock is AWS’s managed service for building generative AI apps and agents, with access to hundreds of foundation models and the AgentCore agent platform.